Timelapse cell tracking

[Mark Aronson]

Hi Oufti Team,

First off, I want to say thank you for building such a wonderful and feature-rich piece of software. I am very excited to get a better handle on using the software for my data analysis. 

In my initial attempts at using Oufti for timelapse imaging of bacterial monolayers, I have been finding the algorithm has been struggling to "keep up with cells" as they grow. For example, these two cells here are well-contoured in this frame of my timelapse:

But then the contour starts to be off on the next frame

And then quite off on the frame after

I have tried changing the following parameters to adjust for this:

• decreased wShedNum to 2000
• increased attrCoeff to 1
• increased neighRep to 12
• increased moveall to 0.5
• increased roiBorder to 50

But nothing seems to be fixing the problem. Any ideas for other parameters to tweak or strategies to make sure the contour properly grows with the cells?

Thank you,

Mark

[Oufti]

Hi Mark,

Thanks for using Oufti!

My recommendation would be to review the parameters and their descriptions in the Oufti web page and focus on the segmentation module to get the best segmented image possible. From there, it might be most helpful to play with the parameters on the provided example data sets and see how the cell meshes are affected. In general, I would adjust one parameter at a time and go from there.

At quick glance, it looks to me that the attrCoeff should be increased and the neighRep decreased, but unfortunately we are unable to tune parameters for all users. Another option would be to refine these cell meshes (by clicking the refine button) and to restart the time-lapse analysis from that specific frame.

Oufti

Op vrijdag 22 januari 2021 om 22:19:09 UTC+1 schreef marksa...@gmail.com: